A thread-based Python implementation of the LabThings API structure, based on the Flask microframework.
pip install labthings
This example assumes a PretendSpectrometer class, which already has data and integration_time attributes, as well as an average_data(n) method. LabThings allows you to easily convert this existing instrument control code into a fully documented, standardised web API complete with auto-discovery and automatic background task threading.
#!/usr/bin/env pythonimporttimefromlabthingsimportActionView, PropertyView, create_app, fields, find_component, opfromlabthings.example_componentsimportPretendSpectrometerfromlabthings.jsonimportencode_json"""Class for our lab component functionality. This could include serial communication,equipment API calls, network requests, or a "virtual" device as seen here.""""""Create a view to view and change our integration_time value,and register is as a Thing property"""# Wrap in a semantic annotation to automatically set schema and argsclassDenoiseProperty(PropertyView):
"""Value of integration_time"""schema=fields.Int(required=True, minimum=100, maximum=500)
semtype="LevelProperty"@op.readpropertydefget(self):
# When a GET request is made, we'll find our attached componentmy_component=find_component("org.labthings.example.mycomponent")
returnmy_component.integration_time@op.writepropertydefput(self, new_property_value):
# Find our attached componentmy_component=find_component("org.labthings.example.mycomponent")
# Apply the new valuemy_component.integration_time=new_property_valuereturnmy_component.integration_time"""Create a view to quickly get some noisy data, and register is as a Thing property"""classQuickDataProperty(PropertyView):
"""Show the current data value"""# Marshal the response as a list of floatsschema=fields.List(fields.Float())
@op.readpropertydefget(self):
# Find our attached componentmy_component=find_component("org.labthings.example.mycomponent")
returnmy_component.data"""Create a view to start an averaged measurement, and register is as a Thing action"""classMeasurementAction(ActionView):
# Expect JSON parameters in the request body.# Pass to post function as dictionary argument.args= {
"averages": fields.Integer(
missing=20, example=20, description="Number of data sets to average over",
)
}
# Marshal the response as a list of numbersschema=fields.List(fields.Number)
# Main function to handle POST requests@op.invokeactiondefpost(self, args):
"""Start an averaged measurement"""# Find our attached componentmy_component=find_component("org.labthings.example.mycomponent")
# Get arguments and start a background taskn_averages=args.get("averages")
# Return the task informationreturnmy_component.average_data(n_averages)
# Create LabThings Flask appapp, labthing=create_app(
__name__,
title="My Lab Device API",
description="Test LabThing-based API",
version="0.1.0",
)
# Attach an instance of our component# Usually a Python object controlling some piece of hardwaremy_spectrometer=PretendSpectrometer()
labthing.add_component(my_spectrometer, "org.labthings.example.mycomponent")
# Add routes for the API views we createdlabthing.add_view(DenoiseProperty, "/integration_time")
labthing.add_view(QuickDataProperty, "/quick-data")
labthing.add_view(MeasurementAction, "/actions/measure")
# Start the appif__name__=="__main__":
fromlabthingsimportServerServer(app).run()Much of the code surrounding default response formatting has been liberally taken from Flask-RESTful. The integrated Marshmallow support was inspired by Flask-Marshmallow and Flask-ApiSpec.
npm install -g conventional-changelog-clinpx conventional-changelog -r 1 --config ./changelog.config.js -i CHANGELOG.md -s